夏赵建+杜友福



摘要:为了解决由于葡萄酒的组成成分十分复杂与多样而导致它的类型分类与品质划分困难的问题,减少由于感官评审法对葡萄酒的评测周期长、影响因素多,主观性强和重复性差以及近红外光谱和三维荧光光谱等重复性差、人的感官长时间易出现适应性疲劳而导致对葡萄酒的口感、质量评测造成不可估计的经济影响。通过对178个葡萄酒样品化学分析数据进行分析处理, 其中有13个葡萄酒属性,使用朴素贝叶斯分类算法对其属性进行分析来确定葡萄酒品质的分类。
关键词:葡萄酒分类;葡萄酒评测;朴素贝叶斯;数据集;葡萄酒样品
中图分类号:TP399 文献标识码:A 文章编号:1009-3044(2017)29-0224-03
Abstract: In order to solve the problem of type classification and difficulty in quality division due to the composition of the wine is very complex and diverse, reduced due to the sensory evaluation method for wine evaluation period is long, the influencing factors are too strong, the subjectivity is strong and the repeatability is poor and the near infrared spectrum and the three-dimensional fluorescence spectrum are reproducible, the human senses are prone to adaptive fatigue for a long time and lead to the taste of the wine , Quality assessment results in unpredictable economic impact. The chemical analysis data of 178 wine samples were analyzed and analyzed, among which there were 13 wine attributes, and the properties of wine were analyzed by using naive Bayesian classification algorithm to analyze the classification of wine quality.
Key words:Classification of wines; Wine evaluation; Naive Bayesian; data set; Sample of wine
1 概述
隨着经济的发展和生活质量的提高,葡萄酒越来越受到大家的欢迎,它的产量和需求量迅速增长,但是它在不适宜的环境中保存容易变质,不仅保存的技术需要复杂的化学知识,造成繁琐、 费时和费用较高, 而且不能对它的组成成分进行全面的分析[1]。感官评测的影响因素多、 主观性比较强,没有统一的参考标准,其重复性差是一个很严重的问题,这种方法对葡萄酒的质量评测及分类造成了难以估量的经济影响[2-3]。葡萄酒的口感与质量是它组成成分的综合反映,由于葡萄酒的组成成分十分的复杂与多样,所以对葡萄酒的类型分类与品质划分比较困难。采用朴素贝叶斯根据酒的13种成分来对酒的种类进行分类。
2 实验数据与朴素贝叶斯的应用……p>